论文部分内容阅读
海面舰船目标的检测与识别对于海面监测与目标打击具有重要意义,弱小舰船目标由于缺少纹理信息且易受到海面阴影、噪声等因素的影响,使得目前常用的检测方法效果较差。基于通道分离与负值扩展对比敏感函数提出了的海面弱小舰船目标检测方法。该方法首先构建像素强度通道与噪声-边缘通道的多分辨图像尺度金字塔;之后,构建不同尺度空间下的负值扩展对比敏感度函数,调制对应各位置的权重;最后,利用各空间尺度系数加权获得两通道视觉显著性图像,通过通道差分处理实现了含噪图像中弱小舰船目标的快速检测。实验结果表明:与其他5种算法相比较,提出的方法具有较高的检测准确率(97.30%)、召回率(84.71%)及综合评价指标(94.49%),同时具备较强的抗噪声能力,适用于含噪海面光学遥感图像中弱小舰船的检测。
The detection and recognition of targets on the surface of the sea are of great significance for the monitoring of the sea surface and target strike. Due to the lack of texture information and vulnerability to the influence of the sea surface shadow and noise, the weak target of the warship makes the commonly used detection methods less effective. Based on the channel separation and the negative extended contrast sensitivity function, this paper proposes a target detection method for weak warship on the sea surface. Firstly, a multi-resolution image scale pyramid with pixel-intensity channel and noise-edge channel is constructed. Then, a negative contrast expansion function under different scales is constructed to modulate the weight of each position. Finally, the spatial scale coefficients are weighted Obtained two-channel visual saliency image, through the channel differential processing to achieve a weak image in the noisy ship targets fast detection. The experimental results show that compared with the other five algorithms, the proposed method has high detection accuracy (97.30%), recall rate (84.71%) and comprehensive evaluation index (94.49%), and has strong anti-noise ability , Suitable for the detection of weak warships in the noisy sea surface optical remote sensing image.